U.S.

Department of Commerce

Bureau of Standards

RESEARCH PAPER RP669

Part of Bureau of Standards Journal of Research,

vol. 12,

April 1934

THERMAL CONDUCTIVITY OF SOME IRONS AND STEELS

OVER THE TEMPERATURE RANGE 100 TO 500 C
By

S.

M.

Sheiton

abstract
The thermal conductivities over the temperature range 100 to 500 C have
been determined for 20 irons and steels which were selected as typical examples
of commercial materials used for a variety of purposes and expected to have
considerably different thermal conductivities.
The data on the chromium-iron
anc chromium-nickel-iron alloys are of particular interest because of the lack of
previous data on the thermal conductivity of "stainless" steels.
The apparatus was designed for comparative measurements and thus eliminated
calorimetric or power-input determinations which are difficult to perform with
uniformly high accuracy over a broad range of temperature. High-purity lead
was used, either directly or indirectly, as the standard with which other metals
were compared.
The

results indicate that, in general, the differences in conductivity of irons

steels are much smaller at high temperatures than at room temperatures.
High-alloy steels have lower thermal conductivities than low-alloy steels.
The
thermal conductivities of irons and low-alloy steels decrease with increase in
temperature.
The conductivities of the high-alloy steels increase with increase
in temperature; in other words, an increase in the amount of alloying constituents in iron causes, in general, a decrease in thermal conductivity and an increase
in the temperature coefficient.
The many and sometimes conflicting factors
concerned make it practically impossible to generalize on the quantitative relationship of thermal conductivity and total alloy content of ferrous metals.
fairly complete bibliography of data on the thermal conductivity of iron and
steel is given.

and

CONTENTS
Page

Introduction
II. Test method
III. Materials and results
IV. Discussion
V. Bibliography
I.

_ _

I.

441
442
442
448
449

INTRODUCTION

Previous determinations of the thermal conductivities of iron alloys

are surprisingly few.
Hall x gives an excellent summary of the
investigations made by Forbes, Lorenz, Angstrom, Neumann and
others during the nineteenth century.
fairly complete bibliography of subsequent data relating to the thermal conductivity- of
iron and its alloys is appended to this paper.
The primary object of most of the earlier researches was to correlate
thermal and electrical conductivities. The test material was of

secondary importance.

Later

tests

were made

on present day

structural materials and the results are more reliable and have more
practical application.
Information on the thermal conductivity of
1

E. H. Hail, concerning thermal conductivity of iron, Physical Keview, vol.

10, p. 277, 1900.

441

442

Bureau

of Standards Journal oj Research

\voi. is

structural materials over a range of temperature is of considerable

2
field
in the selection of materials and in
the design of structures.
In this paper are given results of the determination of the thermal
conductivities of 20 miscellaneous irons and steels which were selected
as typical examples of commercial materials used for a variety of

importance within a limited

purposes and were expected to have considerably different thermal

conductivities.
The data on the chromium-iron and chromiumnickel-iron alloys are of particular interest because of the lack of
previous data on the thermal conductivity of stainless steels.
II.

TEST METHOD

The method and apparatus used in the determinations given in this

paper are described in detail in a separate paper. 3
The apparatus was designed for comparative measurements and
thus eliminated calorimetric or power-input determinations which are
difficult to perform with uniformly high accuracy over a broad range
of temperature. High-purity lead was used, either directly or indirectly, as the " standard" with which other metals were compared.
The thermal conductivity of lead at ordinary temperatures is believed
to be known with as high an accuracy as that of any metal.
The method consisted essentially in measuring the axial temperature gradients in two cylindrical bars, soldered together end to end.
One end of the system was heated, and the other was cooled. The
convex surface of the bars was protected from loss or gain of heat by a
shield tube.
After a steady state of heat flow has been attained, the
heat flux is the same in both bars, and the conductivity at any point
in either bar is inversely proportional to the temperature gradient at
that point.
If the absolute value of the conductivity of the metal
of one bar is known at some temperature within the experimental
range, the conductivity of the other metal can be calculated at all
points where the temperature gradient has been measured.
III.

The

MATERIALS AND RESULTS

made

materials on which determinations were

are described

in table 1.
The results of thermal conductivity determinations of the irons and
low-alloy steels are shown graphically in figure 1
The results on the
.

high-alloy steels, including chromium steels and a high-manganese

nickel steel, are shown in figure 2, and those on the "eigh teen-eight"
type chromium-nickel steels in figure 3
Thermal conductivity values
for different temperatures, interpolated from the experimental results, are listed in table 2.
The value for the thermal conductivity
-1
-1
of lead, 0.352 watts cm
deg at C (International Critical Tables,
vol. 5, page 218) was used as the basis for all determinations.
.

2 M.
S. Van Dusen, note on applications of data on the thermal conductivity of metals.
Effect of Temperature on the Properties of Metals, A.S.M.E.-A.S.T.M., p. 725, 1931.
3 M. S. Van Dusen and S. M. Shelton, B.S.Jour. Research, vol. 12 (RP 668), p. 439, 1934.

Symposium on

Thermal Conductivity of Irons and

Shelton]

d
J
3 _ PI

o^a

.o^g

is

Q O

"C

pq

^^
!

<d

f-

pi

.?a^^-

^wE 8*

Svlfa sl^S: S a*

2"

*'3 ~r^ S S'S m5 '3 s,

-g fl5 a 8 a

800

a m=

?a

""

|.S

4J

"IM

PI

"*

a .S

3M

<

+J '3
bfl fl

ca

2 .^

^>>a
" '"'*

5 o

PI

^a

pq

>>^ a

SOOT'S

a322>Pl*_S*'

-a*.

03

T3

a 85
,)<<

all gSWS
II lip
lllllllllll
l9
rn

LPJ

OT3

I'll

ioT3o3,_,2o0'S<:3o

2" =?

c3

<D c3

TJ.2^S^.2-3fl
dl* 8 l* 8a **

? u

C a o

iil^a

443

Steels

t3

3^3

05

fl

o-n

oo-g|Q
SwO
.2 00 3 .a o

2n
rr-t

pq

<=>

444

Bureau
to at

at

of Standards Journal of Research

at

to at

at

at

to at
An-

<:

Cog-

Cog-

poured

rolled

poured

Rolled

Annealed

poured

Annealed

ingot.

Annealed

ingot.

then

bar.

900

to

round

nch

s,

cooled,

in

s,

ted

ted

%i<z%
a d"3

td

Pi d 3oo
S 3 8 3 t,

S -S f r 31 a-gs
Sa.s<flaa3acBSsaaS35sS-p:8lJ a .a

pq

" .a 3
rA

o3

u-a^-

S S ** 3

03

T3

S 3S-2^1 ^
co

ffl

nd

io

cooled

wn.

ized

o .g*2 .g2 .gi2 o |>~ ^Sg-doo

g .l^-g i^-s i-s % .l^^-gs-sgi

(r

rnace

eated

hen

.s

2 |.2-d-2. 2 8'd-S.HS'

A^as
ictivity

8g|J 1|| fS||

giC g** %%%Z -a lBlii-g .
* |g- !> * s grfSlSa^ajS'gjs?
S

bar.

to

in

water.

d
h4
g^S-a
W
|| llglS 'o W 1B
ace.

to

to

received.

ted

steel

nch

s,

ingot.

then

steel

nch

heated

1120

735
diatoma-

ingot.

then

steel

Andeter-

1120

735

to at

>j

C,

diato(Chemical

rolled
Poured

cc"

C,

Pre-

1-inch

Cog-

C.
rolled

[Vol. It

"'

o-Sta^p

^-gs5.a*"i

-<

{
:

*~i if

o
^y

>o

i-h

S-l

SO

CI

.-H

tH

*
rH

CS

IC

CO

00

<M

lO

:>h

OQ

CO

O
CM

OS

t~

iO

CO

CO

T-H

CO

OS

0>

CN

00

!>

!>

35

t-

CM

i-H

-<*

ui

oo

o
o

o
o

CO

-H

CO

l>

CC

&

o
co

3
o

c
Ph

co

CO

"*

"3

TJ

t--

!>

i-<

-*

oo

oo

a>

-#

t--

0J

CD

CO

co

eo

^a

4>

0>

CO

CM

.-I

i-

'

o"

d
o
d
&H

4)

a
3
d

5
1

-a

-a

'a

b
c

a>

"3

-3

4S

.3

CO

CO

CO

"E

a
3
a

a
3
a

a
3
a

,3

E
C

CD

.2

(-CD
ft

Xi

43

PC

fl-s

CO

CO

CO

4*

a
*

a
a
3
a
o

5
o

O Q

a
a

Xi

Xi

3
a
o

'3

^
a
a

a
o

XJ

J3

o o

CO

d
a

3^
||

u 3
-3

.SP
cot;

Q-

<!
-i

'

<

-<

<

>

t'

oo

=J

<J

fj

<!

<

<

<J

<c


Thermal Conductivity of Irons and

SheUon]

700

Figure
C\

1.

200

JOO
Temperature

S\

C.

-open hearth iron; approximately 99.9 percent iron.

and SiQ

carbon

steel; 0.83

500

400
,

Thermal conductivity of irons and low alloy steels

degrees Centigrade {212 to 1,020 degrees Fahrenheit).

Crwrought iron; approximately 99.5 percent iron.

Cz and C*cast irons; approximately 4.0 percent carbon and

445

Steels

1.5

percent silicon.

percent carbon.

S2 nickel steel; 0.35 percent carbon; 1.37 percent nickel; 0.46 percent chromium.
S3manganese steel; 0.51 percent carbon; 1.65 percent manganese.
Si tungsten steel; 0.35 percent carbon; 1.04 percent tungsten.
S5 chromium steel; 0.10 percent carbon; 5.15 percent chromium.

from 100

to

550

Bureau

446

700

Figure

2.

Thermal

of Standards Journal oj Research

200

300
Temperature

400
,

500

Vol. IS

600

C.

conductivity of high alloy steels from 100 to 600 degrees Centito 1,110 degrees Fahrenheit)

grade (212

A\ chromium steel; 0.08 percent carbon; 15.19 percent chromium.

At- chromium steel; 0.07 percent carbon; 12.0 percent chromium.
A% chromium steel; 0.14 percent carbon; 14.60 percent chromium.
Ai chromium steel; 0.10 percent carbon; 26.00 percent chromium.
Ah chromium-aluminum steel; 1.10 percent carbon; 1.55 percent aluminum; 17.12 percent chromium.
Aimanganese-nickel steel; 0.7 to 0.8 percent carbon; 12 to 13 percent manganese; 3.0 percent nickel.

Antitanium bearing "18-8" stainless; 0.07 percent carbon;

0.34 percent titanium.

Thermal Conductivity of Irons and

Shelton]

447

Steels

0.225

0.200

0.175

!
0.150

!
200

700

300
400
Temperature C.

500

60O

Figure

3.

Thermal conductivity of "18-8"

Centigrade {212

to

stainless steels

from 100

to

550 degrees

1,020 degrees Fahrenheit)

A-,
0.07 percent carbon; 9.10 percent nickel; 18.6 percent chromium.
-480.11 percent carbon; 9.21 percent nickel; 18.5 percent chromium.
4g0.24 percent carbon; 8.96 percent nickel; 19.6 percent chromium.
4io and A\oA 0.24 percent carbon; 7.99 percent nickel; 19.6 percent chromium.

Table

2.

Thermal conductivity, watts cm~ l deg~ x interpolated for

,

listed

Designation

from

Material

100

Basic open-hearth iron

Ci..
Ci__

Wrought

Ci.

iron

Cast iron
Cast iron
Plain carbon

c 4 ...
Si...

steel

SiQ.

Same, quenched

S2

Low-nickel

Si...

Low-manganese steel
Tungsten steel
Low-chromium steel

S3
Si.-

Chromium
Chromium
Chromium

Ai._
Ai._.

A ...
A4
As
3

steel

steel
steel
steel

High-chromium

steel

Chromium-aluminum

steel

Manganese-nickel steel
Chromium-nickel steel
Chromium-nickel steel - ^
Chromium-nickel steel .
Chromium-nickel steel
Same, annealed
Chromium-nickel-titanium

A._.

At
A ._
8

A...

Am-.
AioA
Aii-

The values

the temperatures

the experimental results

steel

0.665
.589
.550
.528
.458
.412
.445
.403
.385
.366
.261
.249
.243
.209
.177
.148
.164
.163
.156
.149
.150
.161

200

0.607
.543
.519
.502
.435
.408
.427
.389
.371
.358
.262
.259
.247
.219
.188
.160
.177
.176
.172
.164
.166
.176

300

0.549
.497
.488
.476
.413
.396
.409
.376
.363
.351
.262
.268
.252
.229
.199
.171
.190
.189
.187
.178
.181
.191

400

0.491
.451
.458
.449
.390
.381
.391
.363
.352
.343
.262
.277
.256
.238
.210
.183
.203
.201
.202
.193
.196
.206

500

0.435
.405
.426
.422
.367
.364
.372
.349
.341
.336
.263
".'261

.243
.221
.195
.216
.214
.217
.207
.212
.221

relative to lead are believed to be accurate within

Experimental data in the literature indicate that the
thermal conductivity of lead at
C is known to an accuracy of 3
percent or better. If at any time the thermal conductivity of the
lead standard is measured by an absolute method and the present
accepted value is found to be in error, the data on all of the materials
tested may be corrected by proportionate adjustment.

2 percent.

4343734-

448

Bureau

of Standards Journal of Research

IV.

[vol. it

DISCUSSION

The results indicate that, in general, the differences in conductivity

of irons and steels are much smaller at high temperatures than at
room temperatures.

High-alloy steels have lower thermal conductiviThe thermal conductivities of iron and
low-alloy steels decrease with increase in temperature.
The conductivities of the high-alloy steels increase with increase in temperature; in other words, an increase in the amount of alloying constituents in iron causes, in general, an increase in the temperature coefficient of thermal conductivity.
The effect on thermal conductivity of small differences in composition is more marked in iron alloys having a small amount of alloying
elements than in alloys with higher alloy concentrations. An additional and somewhat complicating factor is the nature of the alloy.
For example, the total alloy content of cast iron (C 3 ) is greater than
the alloy content of a plain carbon steel (SO but the thermal conductivity at 100 C of the cast iron was found to be approximately 20
percent higher than the conductivity of the plain carbon steel at the
same temperature. The results on the plain carbon steel (S and SiQ)
show that a heat treatment which resulted in a change in structural
constitution produced an appreciable change in conductivity at
lower temperatures. The many, and sometimes conflicting, factors
concerned make it practically impossible to generalize on the quantitative relationship of thermal conductivity and total alloy content
ties

than low-alloy

steels.

of ferrous metals.

With the exception of quenched plain carbon steel (SiQ), the

thermal conductivities of all the materials tested were found to be
linear functions of temperature within the range of temperature in
which measurements were made. An unusual " scatter " is noticeable
in the plotted points representing the determinations on the chromium
The differences are irregular in character and the most
steel (A 2 ).
probable explanation is that a structural change in the specimen
took place with repeated heating in the apparatus. The irregular
character of the "scatter" supports the belief that it cannot be
attributed to experimental error. A typical example of "scatter"
evidently due to experimental error or to a lack of homogeneity of the
specimen is shown by the points representing the determination on
The results of the first series of determinations at relatively
steel A 3
low temperatures are represented by the groups of points numbered
The results of the second series at relatively
1, 2, and 3 (fig. 2).
high temperatures are represented by the groups 4, 5, and 6. The
character of the "scatter" is the same for both series although the
temperatures were quite different. A more complete investigation
over a wider range of temperature is desirable but was not possible
in the present apparatus.
The "18-8" chromium-nickel steels studied represent low-carbon
steels (0.07 to 0.11 percent) in the annealed condition (A 7 ) and in the
quenched condition (A 8 ) and high-carbon steels (0.24 percent) in the
corresponding conditions annealed (A 9 ) and quenched (Ai ). As
shown by the curves in figure 3, the results on the two low-carbon
steels coincided within the experimental error at low temperatures,
and very nearly so at higher temperatures. The higher carbon specimens showed differences in both thermal conductivity and tempera.

sheUon]

Thermal Conductivity of Irons and

Steels

449

greater difference
ture coefficient over the whole temperature range.
In order
in composition existed between the two high-carbon steels.
to determine whether the differences in thermal conductivity of the
high-carbon steels could be attributed to composition or to previous

heat treatment, the quenched specimen was annealed and the thermal
conductivity was redetermined. According to the results, the thermal
conductivity of the reannealed steel was unchanged at low temperature,
but at higher temperatures the annealed material had a higher conIt is noteworthy, and possibly
ductivity than the quenched material.
of some significance, that the thermal conductivity of the annealed
steel was practically the same as that in the quenched state, but the
temperature coefficient of the reannealed specimen A 10 A was of the
same order of magnitude as that of the annealed specimen (A 9 ) of a
different composition.
V.

BIBLIOGRAPHY

E. H. Hall, concerning thermal conductivity of iron.

Physical Review, vol.

10, p. 277, 1900.

W. Jaeger and H. Diesselhorst, Warmeleitung, Electricittasleitung, WarmeWissenschaftliche Abhandlungen

capazitat und Thermokraft einiger Metalle.
der Physikalisch Technischen Reichsanstalt, vol. 3, p. 269, 1900.
E. Griineisen, Uber die Bestimmung des metallischen Warmeleitvermogens und
Annalen der Physik, vol.
iiber sein Verhaltnis zur elektrischen Leitfahigkeit.
308, p. 43, 1900.
Comptes Rendus, vol.
A. Lafay, magnetism and thermal conductivity of iron.
136, p. 1308, 1903.
G. Glage, F. E. Neumann's Methode zur Bestimmung der Warmeleitfahigkeit
gut leitender Korper in Stab- und Ringform und ihre Durchfuhrung an Eisen,
Annalen der Physik,
Stahl, Kupfer, Silber, Blei, Zinn, Zink, Messing, Neusilber.
vol. 323, p. 904, 1905.
C. H. Lees, the effects of temperature and pressure on the thermal conductiviPhilosophical Trans. Rov. Soc. of London, series A, vol. 208,
ties of solids.
p. 381, 1908.
N. F. Smith, the effect of tension on thermal and electrical conductivity.
Pyhsical Review, vol. 28, p. 107, 1909.
U. Bordoni, influence of magnetization on the thermal conductivity of iron and
II Nuovo Cimento, vol. 20, p. 159, 1910.
steel.
C. Hering, the proportioning of electrodes for furnaces.
Trans. Am. Inst.
Elec. Engrs., vol. 29, p. 485, 1910.
T. Simidu, on the thermal and electric conductivities of carbon steels.
Science
Reports of the Tohoku Imperial University, vol. 6, p. Ill, 1917.
K. Honda and T. Simidu, on the thermal and electrical conductivities of carbon
steels at high temperatures.
Science Reports of the Tohoku Imp. Univ., vol. 6,
p. 219 1917.
K. Honda, on the thermal and electric conductivities of nickel steels. Science
Reports of the Tohoku Imp. Univ., vol. 7, p. 59, 1918.
T. Matsushita, on the influence of manganese on the physical properties of
carbon steels. Science Reports of the Tohoku Imp. Univ., vol. 8, p. 79, 1919.
K. Honda, on some physical constants of iron-cobalt allovs. Science Reports,
Tohoku Imp. Univ., vol. 8, p. 51, 1919.
K. Honda and T. Matsushita, on some physical constants of tungsten steels.
Science Reports, Tohoku Imp. Univ., vol. 8, p. 89, 1919.
T. Matsushita, on some physical constants of chromium steels.
Science
Reports, Tohoku Imp. Univ., vol. 9, p. 243, 1920.
L. R. Ingersoll et al, some physical properites of nickel-iron alloys.
Physical
Review, vol. 16, p. 126, 1920.
M. Jakob, Messung des Warmeleitvermogens von Flussigkeiten, Isolierstoffen
und Metallen. Zeitschrift des Vereins deutscher Ingenieure, vol. 66, p. 688, 1922.
E. E. Hall, the variation with temperature of the thermal conductivity of cast
iron.
Physical Review, vol. 19, p. 237, 1922.

450

Bureau

of Standards Journal oj Research

ivoi. 12

J. W. Donaldson, the thermal conductivities of gray cast irons.

Proc. Inst.
Mech. Engrs., London, p. 953, 1928.
P. Graf, Die Warmeleitung von Grauguss.
Giesserei-Zeitung, vol. 26, p. 45,

1929.
J.

W. Donaldson, thermal conductivity

of

metals and alloys.

Mettalurgia,

vol. 1, p. 119, 1930.

Fourth
J. W. Donaldson, the thermal conductivities of ingot mold irons.
Report on the Heterogeneity of Steel Ingots. Supplementary volume, Jour.
Iron & Steel Inst., London, 1932.
W. G. Kannuluik, thermal and electrical conductivity of several metals between
183 and 100 degrees Centigrade, Proc. Royal Society, London, series A, vol. 141,
no. A. 843, p. 159, 1933.

Washington, July

22, 1933.